Faster Whisper Gpu

• Outcome: Pass
• Latest audit: May 28, 2026 at 5:44 AM UTC
• Version:
  0.1.0

Overview

This is a coherent local transcription skill with normal dependency/model downloads and user-chosen transcript output files.

Install in a Python environment where you are comfortable downloading Faster Whisper, PyTorch, and Whisper model files. Prefetch models if you need offline use, and choose output filenames in safe directories to avoid overwriting important files.

SkillSpector

By NVIDIA

Vulnerability Patterns

• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (1)

Description-Behavior Mismatch

Medium

• Confidence: 95% confidence
• Finding: The skill advertises transcription as '100% local' and says no data leaves the machine, but it also states that models are automatically downloaded from Hugging Face on first use. Even if audio content is not uploaded, this still creates external network activity and can violate user expectations, offline-use assumptions, or restricted-environment policies.

VirusTotal

64/64 vendors flagged this skill as clean.